Discrimination of thermodynamic models applied to supercritical-fluid extraction

Experimental planning associated with model discrimination is a technique aimed to optimize experimental procedures causing a cost reduction. Surprisingly, this technique has not been used extensively in phase-equilibrium experiments. Although this rationalization may not be so effective in vapor-liquid equilibrium at low pressures, the same is not true for systems at high pressures. This is the case when considering supercritical-fluid extraction (SCFE) where molecular asymmetry and the proximity to the critical point make modeling of such systems a challenging task. As it has been pointed out in the literature, the success of SCFE relies on the ability to predict and correlate the corresponding thermodynamic behavior. The objective of this work is to show that sequential model discrimination may be used as a rational method for obtaining experimental data, even when a simplified methodology is applied. Literature data on SCFE of two carbon dioxide containing systems were used along with current thermodynamic models. The results show that when model discrimination is applied, less experimental information is required as compared to conventional model selection, without any loss of fitting quality.